Peanut (Arachis hypogaea L.) is widely grown in more than 100

Transcription

1 Research Variability and Stability Analysis for Nutritional Traits in the Mini Core Collection of Peanut Hari D. Upadhyaya,* Ganapati Mukri, Hajisaheb L. Nadaf, and Sube Singh ABSTRACT The nutritional quality of peanut (Arachis hypogaea L.) products depends on the protein content, oil content, and composition of oil. Low genetic variability has been a major bottleneck in genetic enhancement of these nutritional traits in commercial cultivars. The present study was conducted to identify stable genotypes with better nutritional traits and good agronomic performance for use in future breeding programs. The 184 mini core accessions and four control cultivars were evaluated for nutritional traits for two seasons at two locations and for agronomic traits at one location. Significant genotypic and genotype environment interactions were observed for all the nutritional and agronomic traits in the entire mini core collection and within each A. hypogaea subspecies of fastigiata Waldron and hypogaea. Eighteen accessions with higher nutritional traits such as protein content, oil content, oleic acid, and oleic to linoleic acid ratio with superior agronomic traits were identified and their stability analysis resulted in identification of a high oleic acid content (>73%) accession (ICG 2381). On the basis of higher nutritional and agronomic traits 11 subsp. fastigiata and 10 subsp. hypogaea diverse accessions were identified with more than two trait combinations for use in peanut breeding programs for genetic enhancement of nutritional traits. H.D. Upadhyaya and S. Singh, International Crops Research Institute for the Semi-Arid tropics (ICRISAT), Patancheru PO, Hyderabad, Andhra Pradesh , India; G. Mukri and H.L. Nadaf, Dep. of Genetics & Plant Breeding, Univ. of Agricultural Sciences, Dharwad, Karnataka , India. Received 4 May *Corresponding author (h.upadhyaya@cgiar.org). Abbreviations: s 2 g, variance due to genotype; s 2 ge, genotype environment interaction; s 2 gl, genotype location interaction; s 2 gs, genotype season interaction;, linear response; MPLS, modified partial least square; NIR, near-infrared; NIRS, near-infrared spectroscopy; O:L, oleic to linoleic acid ratio; PC, principal component; R, reflectance; S 2 d, nonlinear response; SECV, standard error of cross-validation; UAS, University of Agricultural Sciences; VR, ratio of unexplained variance to total variance. Peanut (Arachis hypogaea L.) is widely grown in more than 100 countries of tropical, subtropical, and warm temperate regions of the globe. This crop is cultivated annually on about million ha worldwide with annual production of million t in shell and productivity of about 1.52 t ha 1 (FAO, 2009). Peanuts are primarily used for extraction of edible oil in Southeast Asian countries and as a food crop with various confectionery and other manufactured products in European and American countries. In recent years, there has been a major shift toward its use as a food crop rather than a source of edible oil in developing countries, including China and India, where protein from animal sources is not within the reach of majority of the population. The quality of the edible oil and various food products of peanut depends on the total oil and protein content in seeds and fatty acid composition of the commercially grown cultivars. Most of the cultivars (with exceptions of high oleic mutants and cultivars bred for Published in Crop Sci. 52 (2012). doi: /cropsci Published online 4 Oct Crop Science Society of America 5585 Guilford Rd., Madison, WI USA All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

2 high oleic trait) and germplasm belonging to A. hypogaea L. subsp. fastigiata Waldron have almost equal proportions (40% each) of oleic acid and linoleic acid while cultivars of A. hypogaea L. subsp. hypogaea possess slightly higher (45 60%) oleic acid compared to linoleic acid (25 30%). These two fatty acids together with a saturated palmitic acid, constitute the major bulk (>90%) of fatty acids in peanut. Among these major fatty acids, oleic monounsaturated fatty acid has the dual advantage of enhanced shelf life of the products and health benefits with lower cardiovascular and other related health risks (Frankel, 1991; Carlson, 1995; Fraser et al., 1997). Polyunsaturated fatty acids have a high potential of developing off flavors during storage due to high oxidation. The rate of oxidation of linoleic acid has been reported to be 10 times higher compared to oleic acid (Frankel, 1991; Lands, 1997); therefore, products of higher oleic to linoleic acid ratio (O:L) have a longer shelf life. The research results of Mozingo et al. (2004) showed a significant advantage of high oleic peanuts for extending shelf life of large-seeded, Virginia-type peanuts for either roasted or salted in-shell processing. Gorbet and Knauft (1997) found that products of SunOleic 95R, a high oleic runner cultivar, had a much longer shelf life than those from the traditional runner-type peanut cultivars. Both saturated fatty acids and polyunsaturated fatty acids in high proportions are harmful to health due to their hyper or hypocholesterolemic nature whereas monounsaturated oleic acid has been reported to be neutral (Groff et al., 1996; Grande and Denke, 1990). Oleic acid lowers harmful low-density lipoprotein (LDL) as effectively as linoleic acid but does not affect beneficial high-density lipoprotein (HDL) levels in blood serum (O Bryne et al., 1997; Kris-Etherton, 2001). Therefore, high oleic peanut cultivars with high oil content for oil extraction and high oleic cultivars with high protein content for peanut products are preferred, but the efforts to breed for such cultivars are lacking especially in developing countries due to insufficient genetic variability for the traits (Norden et al., 1987). The evaluation and screening of germplasm and wild species collections in the United States have indicated low genetic variability for oil and protein content and fatty acid composition (Hammonds et al., 1997). A total collection of 15,445 accessions representing global germplasm are maintained at ICRISAT in Patancheru, India. When the size of the collection is too large, it becomes unmanageable to evaluate and identify useful germplasm. A very small proportion of the total accessions are being used in peanut breeding programs. Alternatively, after constructing a core collection (1704 accessions) of the world germplasm collection (Upadhyaya et al., 2003), an even smaller mini core collection (184 accessions) was developed using taxonomical, geographical, and morphological descriptors (Upadhyaya et al., 2002). The mini core collection represented variability present in the core collection that represented variability in the entire germplasm collection. Evaluation of the mini core collection has resulted in identification of diverse accessions with drought tolerance (Upadhyaya, 2005), salinity tolerance (Srivastava, 2010), high yield potential, high shelling percentage, 100-seed weight (H. Upadhyaya, unpublished data, 2009), and multiple resistance to diseases (Kusuma et al., 2007). There have been limited attempts to evaluate diverse germplasm for nutritional traits in combination with agronomic performance. Our objective was to identify diverse accessions in the mini core collection with high oil content, protein content, and high oleic acid with high O:L coupled with better agronomic performance for use in breeding enhanced nutritional traits in peanut. MATERIALS AND METHODS The peanut mini core collection, consisting of 184 germplasm accessions selected from 1704 core collection accessions, representing the global germplasm of 14,310 accessions maintained at ICRI- SAT genebank (Upadhyaya et al., 2002) was utilized in the present investigation. The mini core collection consisted of 37 subsp. fastigiata var. fastigiata, 58 subsp. fastigiata var. vulgaris Harz, two subsp. fastigiata var. peruviana Krapov. & W. C. Greg., and one subsp. fastigiata var. aequatoriana Krapov. & W. C. Greg. botanical varieties belonging to subsp. fastigiata and 85 subsp. hypogaea var. hypogaea and one subsp. hypogaea var. hirsuta J. Kohler. The four control cultivars used in the study were M 13, ICGS 76 (both subsp. hypogaea var. hypogaea), ICGS 44 (subsp. fastigiata var. vulgaris), and Gangapuri (subsp. fastigiata var. fastigiata). M 13 (Reddy, 1988) and Gangapuri (Isleib et al., 1994) are cultivars developed by the national research centers whereas ICGS 44 (ICGV [PI ]) (Nigam et al., 1990) and ICGS 76 (ICGV [PI ]) (Nigam et al., 1991) are high-yielding cultivars developed at ICRISAT and released for cultivation in India. These 188 genotypes were evaluated in an a design (four blocks each of 47 plots) with three replications in the 2008 rainy and 2008/2009 post-rainy seasons at ICRISAT- Patancheru in alfisol and at the University of Agricultural Sciences (UAS), Dharwad, India, in vertisol fields. These geographical locations, ICRISAT-Patancheru at E, N, and 536 m altitude and Dharwad at E, N, and 678 m altitude and two distinct seasons, rainy and post-rainy, represented diverse environments for investigating the variation and stability of the mini core collection for the nutritional traits. Each plot consisted of a single 4-m row with 60 cm spacing between rows at ICRISAT and 45 cm at UAS, Dharwad. The distance was 10 cm between the plants within each row. The seeds were planted at uniform depth. All required agronomic practices and plant protection measures against pests and diseases to raise a successful crop were followed. The experiments received 60 kg N, 90 kg P 2 O 5, and 60 kg K 2 O ha 1 at UAS, Dharwad, with 10 irrigations in the post-rainy season and no irrigation (rainfed condition) during the rainy season. While at ICRISAT, the experiments received 60 kg P 2 O 5 and 400 kg gypsum ha 1 with 12 irrigations in the post-rainy and six irrigations in the rainy season, each irrigation totaling 5 cm of water. In each plot, five representative plants were tagged randomly to record various qualitative and quantitative traits. Data were recorded on a plot basis for days to 50% flowering (days after emergence to the stage when 50% of the plants had begun flowering), pod yield (kilograms

3 per hectare), shelling percentage, and 100-seed weight (grams). The entire plot was harvested and immature pods were removed, air dried, cleaned, and weighed. The yield of five tagged plants was added to determine total plot yield. A 200 g matured pod sample was used to estimate shelling percentage. The seeds and shells were separated and shelling percentage was computed as follows. Shelling percentage = (weight of seeds 100)/ (weight of seeds + weight of shells) A random sample of 100 seeds was used to record 100-seed weight and protein content; oil content and fatty acid composition were measured from a random seed sample drawn from each genotype with near-infrared spectroscopy (NIRS) (Panford, 1990; Misra et al., 2000) Spectral Measurement of Near-Infrared Spectroscopy Near-infrared spectroscopy differed reflectance spectra were collected by a monochromatic near-infrared (NIR) spectrometer model 6500 (Foss NIRsystems, Inc., Laurel, MD) with spectral range of 400 to 2500 nm and with a light source of tungsten halogen lamps of 50 W and 12 V. The spectrometer was equipped with a silicon detector. For analysis, the seeds were placed in a special adaptor of 3 mm thickness and 37 mm diameter with a central hole of 6 mm. Spectral scanning was done by inserting the adaptor in a standard ring cup (IH-0325) and a sample was placed in the central hole. Before the spectral acquisition, a reference spectrum was collected from a standard check cell (IH-0324). The instrument diagnostics were performed to test the response of instrument, wavelength, and NIRS repeatability to avoid the effects of the surrounding environment on the instrument performance. The absorbance spectral data was recorded as log of reciprocal of reflectance (log 1/R) from 400 to 2500 nm at 8 nm intervals (Kim et al., 2007). The method permitted the analysis of about 40 samples per hour. Mathematical procedures on the spectral information were performed with the Win ISI II Project Manager Software version 1.5 (Infrasoft International, 2000). Calibration The NIR spectrometer (model 6500, Foss NIRsystems, Inc.) was calibrated using chemical reference methods with the application of multivariate regression models to interpret chemical information encoded in the spectral data. Original reflectance spectra were corrected before calibration by applying first and second derivative information, standard normal variate transformation, and de-trend scatter correlation, and four passes were used to eliminate outliers. A second derivative was calculated from log 1/R spectra of six data points and a smoothing over segments of four data points (2, 6, 4, and 1). This combination was selected after having tested six additional math treatments (not included) with and without spectral correlations and data pretreatments. The four data point treatment was either equal or superior in all cases based on standard error of cross-validation (SECV) and 1 VR (1 minus the ratio of unexplained variance to total variance) statistics. The calibrate equations were developed using principle component regression, partial least square, and modified partial least square (MPLS) regression models. Wavelength at an interval of 8 nm across the entire visible plus infrared spectrum was used for calibrations. The SE of calibration, SECV, r, and 1 VR statistics were used to select the best calibration equation. Validation of Calibration Statistics The performance of the calibrated equation were maintained using the cross-validation and external validation and set of sample (n = 100); SECV, SE of prediction, and r were used to determine the accuracy of prediction. The NIR spectrometer model 6500 (Foss NIRsystems, Inc.) was calibrated to analyze fatty acid composition of peanut using representative samples of both the botanical groups, that is, subsp. hypogaea and subsp. fastigiata. Standard normal variate and de-trend scatter correction with second derivative data pretreatment using the MPLS model was used to derive reliable equations with r = 0.90 (Tillman et al., 2006). Using this equation, from each replication of the treatment given we analyzed five individual seeds from each sample that were of same size and of similar maturity and the average of five consistent scan results were taken for analysis. Significant correlation (r = 0.89) between NIR and gas chromatography values after cross-validation indicated general usage of the calibrated equation for estimation of fatty acid composition of both subsp. fastigiata and subsp. hypogaea groups. The best calibrated equations were used to determine protein content, oil content, and fatty acid composition in seeds of the mini core collection. From the data on eight fatty acid compositions determined, the major components of oleic acid and O:L only were used in the present study, and data on other fatty acids are not given. The quality traits such as protein content, oil content, oleic acid (percent), and O:L were measured on a plot basis. Mean values were used for statistical analysis. Data Analysis Data were analyzed as per an a design following residual maximum likelihood analysis (Patterson and Thompson, 1971) with seasons as fixed and entries as random on GenStat 13.1 (VSN International, 2010). The components of phenotypic variance due to genotype (s 2 g) for the entire mini core and for subsp. fastigiata and hypogaea, genotype environment interaction (s 2 ge) for the entire mini core, and fastigiata and hypogaea and their residual (s 2 e) and SEs were calculated. The significance among environments was tested using Wald (1943) statistics. The stability analysis was performed following Eberhart and Russell (1966) to identify stable accessions for nutritional traits (protein content, oil content, oleic acid content, and O:L). Principal component analysis of data on all traits (13 agronomic and 11 quality) and of the selected 21 entries and control cultivars was performed (Hotelling, 1933). The mean observations for each trait were standardized by subtracting from each observation the mean value of the character and subsequently dividing by its respective SD. This resulted in standardized values for each trait with zero mean and SD of 1. The standardized values were used to perform principal component analysis using GenStat 13.1 (VSN International, 2010). Cluster analysis using the Ward (1963) method was performed with scores of the first five principal components (PCs). RESULTS Residual maximum likelihood analysis of individual environment data indicated significant genotypic variances for

4 Table 1. Estimates of variance from pooled analysis of peanut mini core collection for major nutritional quality and agronomic traits. Source of variation Protein Oil Oleic acid O:L Days to 50% flowering Pod yield (kg ha 1 ) Shelling 100-seed weight (g) s 2 g ** 1.777** ** ** 6.072** ** 4.69** 32.62** s 2 g environment 4.452** 1.452** ** ** 0.534** ** 9.28** 8.93** s 2 fastigiata ** 0.786** ** ** 1.827** ns 7.18** 23.82** s 2 fastigiata environment ** 2.305** ** ** 0.819** ** 4.13* 8.09** s 2 hypogaea 6.588** 0.851** ** ** 1.921** * 4.89* 32.14** s 2 hypogaea environment 6.326** 2.458** ** ** 0.146* ** 5.04* 9.82** s 2 error 3.022** 2.549** 5.571** 0.1** 1.797** ** 29.19** 15.67** *Significant at p = **Significant at p = s 2, variance; s 2 g, variance due to genotype. O:L, oleic to linoleic acid ratio. ns, nonsignificant. all four nutritional (protein content, oil content, oleic acid content, and O:L) and four agronomic (pod yield, days to 50% flowering, shelling percentage, and 100-seed weight [g 100 seeds 1 ]) traits (data not given). The pooled analysis of data collected over four environments revealed significant genotypic variance for all four nutritional traits and four agronomic traits (Table 1) indicating wide variation among accessions for the traits studied in the mini core collection. The genotype environment interaction variance was also significant for all the traits studied indicating the significant role of genotype, season, and location in affecting these traits under diverse environmental conditions. The genotypic variances within the subsp. fastigiata (98) and subsp. hypogaea (86) were significant for all the traits except for pod yield in subsp. fastigiata, indicating significant variation within each of the two subspecies germplasm collection also in the mini core collection. The variance due to interaction of genotype and environment was also significant in both groups of subspecies. The relative magnitude of genotypic variance vis-à-vis genotype environment interaction variance was high (2.5 to 11 times) for most traits and almost equal for oil content indicating greater importance of genotypes whereas pod yield and shelling percentage had more influence from the environment. The mean of the control cultivars in each of the subspecies was considered as a benchmark, and 18 accessions that were significantly superior to the respective control cultivars for the first nutritional trait, protein content, were selected and the same numbers of accessions were selected for the other three nutritional traits. These results along with agronomic performance are presented in Tables 2, 3, and 4. For high protein content (Table 2), 15 subsp. fastigiata accessions were significantly superior (29.11 to 31.29%) to the control cultivar ICGS 44 (24.37%) while three of the subsp. hypogaea accessions were superior (28.94 to 29.22%) to the control cultivar ICGS 76 (23.77%). On average the 15 subsp. fastigiata accessions had 41% more protein than mean protein content of the two subsp. fastigiata control cultivars (21.23) whereas the three subsp. hypogaea accessions had only 19% more protein than the average for the two subsp. hypogaea control cultivars (24.43) (Table 2).These identified accessions, except ICG 36, ICG 1711, and ICG 3746 (which had linear response [ ] values significantly different than unity) and ICG 3421 and ICG 5051 (which had significant nonlinear response [S 2 d]), were found to be stable across the environments for protein content. The agronomic performance for these identified accessions was also good. On average the 15 subsp. fastigiata accessions produced about 5% more pod yield than the average of two subsp. fastigiata control cultivars, and three lines, ICG 36, ICG 2019, ICG 4750 ( kg ha 1 ), yielded similar or better than the best subsp. fastigiata control, ICGS 44. ICG 36 contained 31.29% protein and 5% higher pod yield (2809 kg ha 1 ) and had similar shelling percentage. However, 100-seed weight of these identified accessions was less (31.95 to g) compared to the control cultivar ICGS 44 (49.1 g). In the subsp. hypogaea group, of the three entries identified, two of them, ICG 7963 (2756 kg ha 1 ) and ICG (2729 kg ha 1 ), produced pod yield similar to the subsp. hypogaea control cultivar ICGS 76 (2708 kg ha 1 ). The shelling percentage of these accessions was also similar to the control cultivar. For oil content, 67 mini core accessions were significantly superior to ICGS 44 in the subsp. fastigiata group and three in subsp. hypogaea group had more oil than ICGS 76. The best 18 accessions identified for high oil content from the mini core collection (Table 3) indicated that 15 subsp. fastigiata (48.35 to 50.45%) and three subsp. hypogaea (48.01 to 49.53%) accessions were superior to their respective control cultivars, ICGS 44 (43.74%) and ICGS 76 (46.09%). On average the 15 subsp. fastigiata accessions had about 6.9% more oil than average of two control cultivars (45.78%) and the three subsp. hypogaea accessions had 6.5% more oil than the mean oil content of two control cultivars (45.91). Most of the identified accessions were stable except ICG 397, ICG 3681, and ICG 5779 of subsp. fastigiata (significant S 2 d) and ICG 4746 and ICG of subsp. hypogaea (significant ). The pod yield of these identified accessions was similar (2228 to 2809 kg ha 1 ) in 10 out of 15 accessions in the subsp.

5 Table 2. Performance and stability of top 18 high protein A. hypogaea subsp. fastigiata and subsp. hypogaea accessions selected from the peanut mini core collection. Subspecies and accession Origin Protein S 2 d Days to 50% flowering Pod yield (kg ha 1 ) Shelling 100-seed weight (g) subsp. fastigiata ICG 36 India * ICG 5779 India ICG 3421 India * ICG 3584 India ICG 2019 India ICG 1519 India ICG 4729 China ICG 1973 India ICG 3746 Argentina ** ICG 4750 Paraguay ICG India ICG 4911 Malawi ICG 2106 India ICG 81 Unknown ICG 1711 Bolivia * Gangapuri India ICGS 44 India subsp. hypogaea ICG 7963 United States ICG 5051 United States ** ICG United States ICGS 76 India * M 13 India * Trial mean SE ± LSD (5%) CV Heritability (h 2 ) *Significant at p = **Significant at p = 0.01., linear response. S 2 d, nonlinear response. fastigiata group compared to control cultivar ICGS 44 (2686 kg ha 1 ). ICG 36 (48.53% oil and 2809 kg ha 1 ) and ICG 1142 (48.79% oil and 2709 kg ha 1 ) produced higher pod yield with similar shelling percentage and days to 50% flowering. However, the three subsp. hypogaea accessions yielded (1279 to 1522 kg ha 1 ) lower than ICGS 76 (2708 kg ha 1 ). The major fatty acid determining oil quality is oleic acid content, and 18 top accessions were identified with high oleic acid, of which seven belonged to the subsp. fastigiata (57.55 to 61.88%) and 11 to the subsp. hypogaea (57.63 to 73.36%) group (Table 4). The selected accessions were superior to the respective controls, ICGS 44 (47.85%) and M 13 (54.9%). The seven subsp. fastigiata accessions had on average 33% more oleic acid than the mean of two control cultivars whereas the 11 subsp. hypogaea accessions had 7% more oleic acid than mean of control cultivars. Six subsp. hypogaea accessions (ICG 2381, ICG 6766, ICG 7243, ICG 10185, ICG 12276, and ICG 15190) and all the subsp. fastigiata accessions except ICG 6022 were stable across environments. The high oleic (>60%) subsp. hypogaea accessions (ICG 2381, ICG 10185, ICG 15419, ICG 12276, ICG 7243, and ICG 6766) except the top accession ICG 2381 (73.36% and 2051 kg ha 1 ) produced similar pod yield ( kg ha 1 ). ICG (3712 kg ha 1 ) produced significantly higher pod yield than control cultivar ICGS 76 (2708 kg ha 1 ). The best three subsp. fastigiata accessions (ICG 6022, ICG 12625, and ICG 11088) recorded 60% oleic acid with higher pod yield ( kg ha 1 ). The identification of ICG 2381, a stable accession, with nonsignificant S 2 d (2.86) and ( 0.25) values and high mean oleic acid (73.36%) is a significant finding. Based on the relative proportion of oleic acid to linoleic acid computed as O:L, 18 top accessions were identified from subsp. fastigiata (9 accessions) and subsp. hypogaea (9

6 Table 3. Performance and stability of top 18 high oil A. hypogaea subsp. fastigiata and subsp. hypogaea accessions selected from the peanut mini core collection. Subspecies and accession subsp. fastigiata Origin Oil Days to S 2 d 50% flowering Pod yield (kg ha 1 ) Shelling 100-seed weight (g) ICG 442 United States ICG 397 United States * ICG 5779 India * ICG 4955 India ICG Cameroon ICG 6201 Cuba ICG 1415 Senegal ICG 1142 Benin ICG 3681 United States ** ICG Uganda ICG 36 India ICG 1519 India ICG 3421 India ICG 3673 Korea ICG 115 India Gangapuri India * ICGS 44 India * subsp. hypogaea ICG 4746 Israel * ICG Mali * ICG Uruguay ICGS 76 India M 13 India Trial mean SE ± LSD (5%) CV Heritability (h 2 ) *Significant at p = **Significant at p = 0.01., linear response. S 2 d, nonlinear response. accessions) groups (Table 4). The identified subsp. fastigiata accessions had significantly higher (2.65 to 3.31) O:L compared to the control cultivar ICGS 44 (1.671). All the identified accessions except ICG 5475 and ICG of subsp. fastigiata were stable across the environments. The best three stable accessions (ICG 6022, ICG 1274, and ICG 12625) with 3.0 O:L and good agronomic traits can be utilized in genetic improvement programs. ICG 2381 from subsp. hypogaea has been identified as an outstanding (6.91) and stable accession for O:L and can be a good donor parent in crop improvement programs. Five other accessions (ICG 10185, ICG 7243, ICG 6766, ICG 12276, and ICG 15419) with 3.0 O:L and good agronomic performance and stable across environments have been identified. The 21 accessions identified with two and more combined nutritional traits are listed in Table 5 and depicted in Fig. 1. ICG 36, ICG 1519, ICG 3421, and ICG 5779 recorded higher protein and oil contents. ICG 36 flowered in ~25 d and produced 2809 kg ha 1 pod with 100- seed weight of g. ICG 6022 (61.8% oleic acid and 3.31 O:L) and ICG (60.26% oleic and O:L) recorded higher pod yield (3229 and 3284 kg ha 1, respectively). Among the 10 subsp. hypogaea accessions ICG 2381 had high oleic acid (73.36%) and O:L (6.91). ICG with 61.12% oleic acid and 3.03 O:L recorded significantly higher (3712 kg ha 1 ) pod yield compared to control In case of nutritional traits, O:L was highly correlated with oleic acid content (r = 0.942) whereas protein and oil contents were unrelated to each other (r = 0.002). However, both protein and oil were negatively correlated with oleic acid (protein oleic acid, r = 0.234, and oil oleic acid, r = 0.656) and O:L ratio (protein O:L, r = and oil O: L, r = 0.646). Except for protein content (r = 0.186), none of the nutritional traits were correlated with

7 Table 4. Performance and stability of top 18 high oleic acid and 18 high oleic to linoleic acid ratio (O:L) A. hypogaea subsp. fastigiata and subsp. hypogaea accessions selected from peanut mini core collection. Subspecies and accessions Origin Oleic acid S 2 d O:L S 2 d Days to 50% flowering Pod yield (kg ha 1 ) Shelling % 100-seed weight (g) subsp. fastigiata ICG 6022 Sudan * ICG Ecuador ICG Peru ICG 1274 Indonesia ICG Brazil * ICG 6375 Unknown ICG 5221 Argentina ICG 8106 Peru ICG 5475 Kenya ** Gangapuri India ** ICGS 44 India ** subsp. hypogaea ICG 2381 Brazil ICG United States ICG Ecuador * ICG Bolivia ICG 7243 United States ICG 6766 United States ICG 928 Unknown * 45.90** ICG 3053 India ** ICG Costa Rica ICG India ** ICG 2777 India ** ** ICGS 76 India * M 13 India Trial mean SE ± LSD (5%) CV Heritability (h 2 ) *Significant at p = **Significant at p = 0.01., linear response. S 2 d, nonlinear response. Selected only for O:L. Selected only for oleic acid. pod yield (r = with oil, with oleic acid, and with O:L). While oil content was negatively correlated (r = 0.297), oleic acid (r = 0.440) and O:L (r = 0.434) were positively correlated with 100-seed weight. Principal component analysis, using data of eight traits of 21 selected accessions and four control cultivars, was performed. The first principal component (PC1) explained 48.5% variation, followed by PC2 (22.5% variation), PC3 (13.0% variation), PC4 (5.5% variation), and PC5 (5.1% variation). Cluster analysis based on scores on first five PCs (94.6% variation) delineated the accessions in three clusters (Fig. 2). Accessions were delineated on the basis of high oleic acid (60.7%), high pod yield (2836 kg ha 1 ), and high 100-seed weight (46.8 g) in cluster1, high O:L (3.39) and early flowering (23.7 d) in cluster2, and high protein (26.6%) and high shelling percentage (66.6%) in cluster 3. DISCUSSION The results of evaluation of the mini core collection over two locations (Patancheru and Dharwad) in two seasons (2008 rainy and 2008/2009 post-rainy) indicated the presence of variation for the four nutritional traits protein content, oil content, oleic acid content, and O:L. The magnitude of variation was comparatively high for protein content in subsp. fastigiata accessions while it was high for O:L in subsp. hypogaea, and for two other traits, oil content and oleic acid content, the magnitude of variation was almost similar in both the groups of genotypes. The

8 Table 5. Performance of 21 A. hypogaea subsp. fastigiata and subsp. hypogaea accessions selected for superior performance over mean of control cultivars for two or more nutritional traits from the peanut mini core collection. Subspecies and accession Origin Protein Oil Oleic acid O:L subsp. fastigiata ICG 36 India ICG 5779 India ICG 3421 India ICG 1519 India ICG 1274 Indonesia ICG 5221 Argentina ICG Brazil ICG 6375 Unknown ICG Ecuador ICG 6022 Sudan ICG Peru Gangapuri India ICGS 44 India subsp. hypogaea ICG 3053 India ICG United States ICG 928 Unknown ICG 2777 India ICG 6766 United States ICG Costa Rica ICG 7243 United States ICG Bolivia ICG 2381 Brazil ICG Ecuador ICGS 76 India M 13 India Trial mean SE ± LSD (5%) CV Heritability (h 2 ) O:L, oleic to linoleic acid ratio. significance of s 2 ge was indicative of the effect of environments on genotypes in the expression of these nutritional quality traits. The s 2 ge was further partitioned into genotype location interaction (s 2 gl) and genotype season interaction (s 2 gs). The variances due to s 2 gl and s 2 gs were of similar magnitude for protein content whereas for other three traits (oil, oleic acid, and O:L), s 2 gl was greater than s 2 gs. Significant genotypic interactions with growing season and geographic location have been reported earlier for fatty acid profiles (Holaday and Pearson, 1974, Norden et al., 1987). Considering proportion of variance due to genotype environment interaction to the total phenotypic variance (Upadhyaya, 2005), oil content (14.76%) was least stable followed by oleic acid (8.82%) whereas the protein and O:L were the most stable nutritional traits. Superior accessions for each of the four nutritional traits have been identified. The mean protein content of the selected subsp. fastigiata (29.84%) and subsp. hypogaea (29.03%) was much higher compared to control cultivars (22.82%) and the mean of all subsp. fastigiata (24.5%) and subsp. hypogaea (22.9%) accessions in the mini core collection. The subsp. fastigiata accessions identified have higher protein content, 5 to 7%, with similar pod yield as compared to high yielding control cultivars and will provide an opportunity for breeders to choose genetically diverse accessions for enhancing protein content in confectionary cultivars. ICG 442, among subsp. fastigiata group, recorded 50% mean oil content. The mean oil content of the selected accessions was greater (48.8%) compared to control cultivars (45.8%). Although high oleic acid with >70% germplasm lines were identified earlier (Norden et al., 1987), they are not available for global research and development. The present study resulted in identification of the stable and high oleic (73.36%) subsp. hypogaea accession (ICG 2381) with similar 100-seed weight (50.8 g) to the control cultivars. This could be the valuable source for peanut breeders for genetic enhancement of oleic acid and to enhance the shelf life of peanut products. The other five accessions (ICG 10185, ICG 15419, ICG 12276, ICG 7243, and ICG 11088) with 60% oleic acid and with either similar or superior performance for agronomic traits could be the additional resources for recombinant breeding. The O:L of these identified lines was also higher indicating their superiority over the control cultivars. The selected accessions of both subsp. hypogaea (60.85%) and subsp. fastigiata (59.51%) had higher oleic acid content compared to the mean of all subsp. hypogaea (52.9%) and subsp. fastigiata (44.5%) accessions and control cultivars (50.78%). The correlations among nutrient and agronomic traits indicated that simultaneous selection for various nutritional and agronomic traits, such as high protein and high pod yield and high oleic acid content and O:L with high 100- seed weight was possible in the mini core collection. Also, the selection for protein did not adversely affect oil content. Twenty-one diverse accessions (11 subsp. fastigiata and 10 subsp. hypogaea) with two or more desired trait combinations have been identified (Table 5). The agronomic performance of the selected accessions for four nutritional traits is graphically represented in Fig. 1. Among these, subsp. fastigiata accessions with combined traits of both high protein content and high oil content or high oleic acid and high O:L were identified whereas in subsp. hypogaea only high oleic acid and high O:L accessions were identified. The identified accessions were geographically as well as genotypically diverse (Fig. 2). Interestingly, three of the control cultivars, one subsp. fastigiata (ICGS 44) and two subsp. hypogaea (M 13 and ICGS 76) were grouped in the same cluster and the identified accessions were found to be diverse from these genotypes. The genetic enhancement of these nutritional traits could be easy as they have higher heritability (>80%).

9 Figure 1. Graphical representation of performance of superior accessions selected for high protein content, oil content, oleic acid content, and oleic to linoleic acid ratio (O:L) and combined nutritional traits in the mini core collection of peanut.

10 Figure 2. Dendrogram of the selected 21 superior peanut mini core accessions and four control cultivars based on scores of first five principal component (94.6% variation). Several lines identified for nutritional traits in this study (Tables 2 through 5) have also been reported as resistant or tolerant to important abiotic (drought, salinity) and biotic (diseases) stresses. The high protein lines, ICG 1519, ICG 1711, and ICG 2106, have also been reported as tolerant to salinity (Srivastava, 2010), ICG 2106 also being tolerant to drought (Upadhyaya, 2005) and ICG 4750 resistant to aflatoxin contamination (Jiang et al., 2010). Similarly, the high oil line ICG 442 has been reported as tolerant to salinity and ICG 4746 and ICG as resistant to rust (Kusuma et al., 2007), high oleic acid lines ICG 7243 as tolerant to drought (Upadhyaya, 2005), ICG as resistant to rust, and ICG as resistant to aflatoxin and late leaf spots (Kusuma et al., 2007), and the lines with high O:L, ICG 6022, as resistant to rust, late leaf spots, and early leaf spots (ICRISAT, 2009) and ICG 6766 as tolerant to drought (Upadhyaya, 2005). The agronomic desirability of these lines is of particular significance in meeting the needs of peanut breeders for traitspecific, diverse, and agronomically desirable germplasm lines for use in their breeding programs to develop peanut cultivars with desirable combinations of traits and with a broad genetic base. Limited quantity of seeds of the lines identified in this study is available for research using Standard Material Transfer Agreement. References Carlson, S.E The role of PUFA in infant nutrition. Inform 6: Eberhart, S.A., and W.A. Russell Stability parameters for comparing varieties. Crop Sci. 6: doi: /cropsci X x FAO FAOSTAT database. Available at org/site/567/default.aspx#ancor (verified 8 Sept. 2011). FAO, Rome, Italy. Frankel, E.N Recent advances in lipid oxidation. J. Sci. Food Agric. 54: doi: /jsfa Fraser, G.E., D. Sumbureru, P. Pribis, R.L. Neil, and M.A. Frankson Association among health habits, risk factors and all-cause mortality in a black California population. Epidemiology 8: doi: / Gorbet, D.W., and D.A. Knauft Registration of SunOleic 95R peanut. Crop Sci. 37:1392. doi: /cropsci X x Grande, S.M., and M. Denke Dietary influences on serum lipids and lipoproteins. J. Lipid Res. 31: Groff, J.L., S.S. Gropper, and S.M. Hunt Lipid. p In Advanced nutrition and human metabolism. West Publishing, Minneapolis/St. Paul, MN. Hammonds, E.G., D. Duvick, D. Wang, D.H. Dodo, and R.N. Pittman Survey of the fatty acid composition of peanut (Arachis hypogaea L.) germplasm and characterization of their epoxy and eicosenoin acids. J. Am. Oil Chem. Soc. 74: doi: /s z Holaday, P.E., and J.L. Pearson Effects of genotype and production area on the fatty acid composition, total oil and protein in peanuts. J. Food Sci. 39: doi: /j tb07355.x

11 Hotelling, H Analysis of a complex of statistical variables into principal components. J. Educ. Psychol. 24: doi: /h ICRISAT ICRISAT archival report Sustaining biodiversity of sorghum, pearl millet, small millets, groundnut, pigeonpea and chickpea for current and future generations. Available at publications/archival-reports/icrisat-arr-2009.pdf (verified 8 Sept. 2011). ICRISAT, Patancheru, India. Infrasoft International The complete software solution using a single screen for routine analysis, robust calibrations, and networking; Manual, FOSS NIRSystems/TECATOR. Infrasoft International, Silver Spring, MD. Isleib, T.G., J.C. Wynne, and S.N. Nigam Groundnut breeding. p In J. Smartt (ed.) The groundnut crop A scientific basis for improvement. Chapman & Hall, London, UK. Jiang, H.F., X.P. Ren, S.Y. Wang; X.J. Zhang, J.Q. Huang, L. BoShou, C.C. Holbrook, and H.D. Upadhyaya Development and evaluation of peanut germplasm with resistance to Aspergillus flavus from core collection. Acta Agron. Sin. 36: doi: /sp.j Kim, K.S., S.H. Park, M.G. Choung, and Y.S. Jang Use of near-infrared spectroscopy for estimating fatty acid composition in intact seeds of rapeseed. J. Crop Sci. Biotech. 10: Kris-Etherton, P.M The effects of nuts on coronary heart disease risk. Nutr. Rev. 59: doi: /j tb06996.x Kusuma, V.P., G. Yugandhar, B.C. Ajay, M.C.V. Gowda, and H.D. Upadhyaya Identification of sources of multiple disease resistance in groundnut (Arachis hypogaea L.) mini core. p In Indian Soc. Oilseeds Res. Natl. Seminar Abstr., Hyderabad, India Jan Directorate of Oilseeds Research, Rajendranagar, Hyderabad, India. Lands, W.E.M The two faces of essential amino acids. Inform 8: Misra, J.B., R.S. Mathur, and D.M. Bhatt Near-infrared transmittance spectroscopy: A potential tool for nondestructive determination of oil content in groundnuts. J. Sci. Food Agric. 80: doi: /(sici) ( )80:23.0.co;2-9 Mozingo, R.W., S.F. O Keefe, T.H. Sanders, and K.W. Hendrix Improving shelf life of roasted and salted inshell peanuts using high oleic fatty acid chemistry. Peanut Sci. 31: doi: /pnut Nigam, S.N., S.L. Dwivedi, Y.L.C. Rao, and R.W. Gibbons Registration of ICGV peanut cultivar. Crop Sci. 30:959. doi: /cropsci x x Nigam, S.N., S.L. Dwivedi, Y.L.C. Rao, and R.W. Gibbons Registration of ICGV peanut. Crop Sci. 31:1096. doi: /cropsci x x Norden, A.J., D.W. Gorbet, D.A. Knauft, and C. Young Variability for oil quality among peanut genotypes in the Florida breeding programme. Peanut Sci. 14:7 11. doi: / i O Bryne, D.J., D.A. Knauft, and R.B. Shireman Lowfat monounsaturated rich diets containing high-oleic peanuts improve serum lipoprotein profiles. Lipids 32: doi: /s y Panford, J.A Determination of oil content of seeds by NIR: Influence of fatty acid composition on wave length selection. J. Am. Oil Chem. Soc. 67: doi: /bf Patterson, H.D., and R. Thompson Recovery of interblock information when block sizes are unequal. Biometriks 58: doi: /biomet/ Reddy, P.S Genetics, breeding and varieties. p In P.S. Reddy (ed.) Groundnut. Publication and Information Division, Indian Council of Agricultural Research, New Delhi, India. Srivastava, N Identification of contrasting parental material for salinity tolerance based on morphological, physiological, and biochemical study and molecular diversity by SSR molecular markers in groundnut (Arachis hypogaea). Ph.D. thesis. Jawaharlal Nehru Technological University, Hyderabad, India. Tillman, B.L., D.W. Gorbet, and G. Person Predicting oleic and linoleic acid content of single peanut seeds using nearinfrared reflectance spectroscopy. Crop Sci. 46: doi: /cropsci Upadhyaya, H.D Variability for drought resistance related traits in mini core collection of peanut. Crop Sci. 45: doi: /cropsci Upadhyaya, H.D., P.J. Bramel, R. Ortiz, and S. Singh Developing a mini core of peanut for utilization of genetic resources. Crop Sci. 42: doi: /cropsci Upadhyaya, H.D., R. Ortiz, P.J. Bramel, and S. Singh Development of groundnut core collection using taxonomical, geographical and morphological descriptors. Genet. Resour. Crop Evol. 50: doi: /a: VSN International GenStat software for windows. Release VSN International Ltd., Hemel Hempstead, UK. Wald, A Test of statistical hypothesis concerning several parameters when the number of observations is large. Trans. Am. Math. Soc. 54: doi: /s Ward, J Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 38: doi: /